52 BOTANY OF SOCOTRA. 



5. Boswellia sp. 



Socotra. On the limestone plains. B.C.S. n. 708. 



Of a small dwarf tree with hard spiny branches, bearing contracted laterals, 

 on which the leaves are clustered, we have specimens. The leaves are unequally 

 pinnate, about § inch long, and there are some seven pairs of sessile leaflets. 

 All the leaflets are tridentate, quite glabrous, and about -fa inch long, and the 

 rhachis is winged. The bark of the tree is dark-coloured, and on the contracted 

 lateral branches the persistent insertions of the leaves give rise to a tuberculate 

 surface, as in the plant I have just described. Can this be the young condition 

 of the foregoing species B. socotrana ? 



2. BALSAMODENDRON. 



Balsamodendron, Kuuth in Ann. Sc. Nat. ser 1, ii. (1824), 348 ; Benth. etHook. Geu. PI. i. 323. 



I have retained Kunth's generic name, which has come into general use, 

 although Baillon (Hist. d. Plantes v. (1874), 310), and after him Engler (Bot. 

 Jahresb. i. (1881), 41), have adopted the name Balsamea of Gleditsch, for which 

 they claim priority. But all evidence proves their assumption erroneous. 

 Gleditsch published in 1782 (Schrift. d. Berl. Gesellsch. Naturf. Freunde 103, t. 

 8, f. 2) a paper entitled " Bemerkung liber das Geschlecht und die Art der achter 

 Balsampflanze von Mecca," in which he describes, under the name Balsamea 

 meccanensis, some dried specimens obtained by Achmet Effendi near Mecca. 

 He contrasts the characters of this plant with those of Amyris gileadensis, Linn. 

 (Mant. 65) — a species described by Linnaeus on specimens sent from Arabia 

 by Forskal — and specially points out how very different the two plants are, 

 that indeed they could not belong to the same genus. The leaves of his 

 plant he describes and figures as bipinnate, the calyx and corolla each of 

 five parts, and larger than in the Linnean plant ; the stamens ten, though 

 they may be eight or nine, exceeding the corolla ; and the immature fruit with a 

 basal pentagonal stipe. He also points out that Linnseus's plant is no Amyris. 

 Now Engler, with whom I have corresponded on this matter, whilst admitting 

 that the floral characters given by Gleditsch are not met with in any of the 

 species of Balsamodendron he has examined, and concluding that " the flowers 

 do not belong to the Burseracea? at all," yet considers the branching and 

 inflorescence agree well with what is observed in Balsamodendron, and the 

 bipinnate leaves he regards as no barrier to such an identification. For my 

 part I fail to recognise in Gleditsch's description and figure the characters of a 

 Balsamodendron. Engler, too, himself shows his position to be untenable, for 

 he identifies Balsamea meccanensis, Gled., as a type of which three Arabian 

 plants, viz. : — Balsamodendron opobalsamum, Kunth (in Ann. Sc. Nat. ser. 1. ii. 

 (1824), 348), Balsamodendron gileadense, Kunth (loc. cit.), and Balsamoden- 

 dron Ehrenbergianum, Berg (in Bot. Zeit. (1862), 163), are merely forms. Now I 



